The QEMU 10.1.0 release has been presented. As an emulator, QEMU allows you to run a program compiled for one hardware platform on a system with a completely different architecture, for example, to run an application for ARM on an x86-compatible PC. In virtualization mode, QEMU's code execution performance in an isolated environment is close to that of a physical system due to direct execution of instructions on the CPU and the involvement of the Xen hypervisor or KVM module in Linux, or the NVMM module in NetBSD.
Originally, the project was created by Fabrice Bellard to enable the execution of Linux executables compiled for the x86 platform on architectures different from x86. Over the years of development, full emulation support for 14 hardware architectures has been added, and the number of emulated hardware devices has exceeded 400. More than 2,700 changes from 226 developers have been made in preparation for version 11.0.
Key improvements added in QEMU 10.1:
- In VFIO (Virtual Function I/O), initial support for memory mapping for confidential guest environments has been implemented, where memory encryption is used to protect against interference and analysis by the host system administrator. This change allows device pass-through to guest systems protected by Intel TDX (Trusted Domain Extensions) and AMD SEV-SNP (Secure Nested Paging) mechanisms.
- The multifd mechanism has been added to speed up Live migration in postcopy mode, where the virtual machine is first transferred and started on the target host, virtual machine, and then the unused memory content is gradually transferred. The Live migration precopy mode has been optimized. RDMA migration support for IPv6 has been added.
- A 'guest-get-load' command has been added in QEMU Guest Agent to retrieve the system load average information in virtual machines running Windows.
- In virtio-gpu, support for name substitution in EDID (Extended Display Identification Data) has been added.
- In the x86 architecture emulator, the Intel TDX (Trusted Domain Extensions) mechanism for encrypting the memory of guest systems running under the KVM hypervisor has been implemented. Support for initializing confidential virtual machines i formatet IGVM (Independent Guest Virtual Machine), der bruger Intel TDX- og AMD SEV-SNP-mekanismer til beskyttelse.
- I ARM-arkitekturens emulator er der implementeret support til boards (Analog Devices âmax78000fthrâ, âast2700fcâ, Meta âcatalina-bmcâ, NVIDIA âgb200-bmcâ og âast2700a0-evbâ. Emuleringen inkluderer udvidelserne af CPU FEAT_SME2, FEAT_SME2p1, FEAT_SME_B16B16, FEAT_SME_F16F16, FEAT_SVE_B16B16 og FEAT_SVE2p1. Den typiske virtuelle ARM-platform âvirtâ har fĂ„et tilfĂžjet support til nested virtualization, nĂ„r KVM bruges, samt support for CXL-bussen (Compute Express Link) og hot-plugging af PCI-enheder ved hjĂŠlp af ACPI.
- I RISC-V-arkitekturens emulator er der implementeret emulering af processoren Kunminghu og platformen baseret pÄ den. Der er tilfÞjet stÞtte til arkitekturens instruktionssÊtsudvidelse Ziccif med implementering af en atomar version af fetch-instruktionen.
- I LoongArch-arkitekturens emulator er der tilfÞjet support til emulering af irqchip-interrupt controllere pÄ kerne-siden.
- Der er tilfĂžjet eksperimentel support til kompilering til WebAssembly (WASM) ved hjĂŠlp af Emscripten-kompilatoren.
- I komponenterne til vĂŠrt-miljĂžer er supporten for Debian 11 ophĂžrt.
Kilde: opennet.ru
